基于初级侧EMF相量估计的IPT系统参数识别及输出电压和电流控制

The performance of a power converter is closely related to the selection of circuit parameters. However, the mainstream parametric design methods at present are overly dependent on humans. It is often not possible to obtain the best combination of key parameters. Therefore, an artificial intelligence-based design method for the key parameters is proposed, which aims to realize the best performance and automated design process. Firstly, a performance estimation model (PEM) for power converters is developed that considers the physical law-attention mechanism (AM). It incorporates AM into a physical information neural network and automatically builds the mapping model of design parameters and objectives. The design objectives represent the different performances of the converter. Secondly, the particle swarm optimization method is used to realize the automated search for the best combination of design parameters. Its optimization ability in discontinuous space is improved by making discrete variables continuous. Finally, the superiority of the proposed method is proved by the comparison experiment and the scalability verification experiment.